The Legal Profession’s AI Tightrope

By James Tuke, AI Futures Forum.

Over the last twelve months, the AI conversation in the legal profession has shifted significantly. In May 2025, we were benchmarking baseline implementation across law firms. By September, we could see corporate clients beginning to harden their efficiency expectations on firms. Then, in January 2026, our Broken Ladder green paper issued a warning: by automating the routine ‘grunt work’ tasks of the profession, we risked hollowing out the talent pipeline.

Today, GenAI is firmly embedded across the profession. Yet, law firms are walking a dangerous AI tightrope. Despite the headline adoption metrics, we are still in the very early stages of this transition, and inevitably, we are very much feeling our way still.

The real challenge ahead is not just a technological one; it is cultural and economic, too. Law firms are engaged in a delicate juggling act, facing intense pressure from corporate clients to balance AI-driven efficiency, reduced costs, and uncompromised quality of service. Managing this internal friction whilst maintaining relentless billable outputs is stretching cognitive bandwidth to its limit. However, the firms that look past short-term gains and get this structural alignment right will be the ones best positioned for the long term.

The Threat of the NewMods

Adding to this pressure is a significant external challenge to traditional partnerships: the rise of the ‘NewMods’. These AI-first, full-stack firms operate without the massive overhead of traditional human pyramids, bypassing the billable hour in favour of transparent, fixed-fee structures.

Traditional firms comfort themselves with the belief that they offer a level of bespoke strategic value that insulates them from such competitors. Whilst that may be true for highly complex cases and ‘bet-the-company’ litigation, firms would be extremely unwise to ignore the NewMod threat. For less complex, highly systematic, high-volume work, these new kids on the block are moving aggressively – with the investment behind them.

Looking Ahead: The Age of the Agent

If Generative AI makes lawyers faster at drafting, the deployment of Agentic AI aims to remove the human from the doing of the work entirely. We are moving from reactive chatbots to autonomous digital workers capable of managing end-to-end legal workflows.

This raises profound questions for practitioners and clients alike. Will we inevitably outsource much, if not all, matters to our digital agents? If so, we may trigger a classic Jevons Paradox: as the transaction cost of accessing legal services plummets, the sheer volume of demand will skyrocket, creating an entirely new ecosystem of machine-to-machine commerce and compliance.

Where, then, does the human remain in the loop? Aside from ultra-complex, high-judgment matters requiring deep strategic empathy, perhaps the answer is simple: we will always need a human in the loop because we need someone to blame – and ultimately sue – should things go wrong. Professional indemnity insurers are already hardening their requirements, demanding explicit AI governance and affirmative cover frameworks before granting renewals. The fiduciary and epistemic liability must ultimately stop with a qualified (human) professional.

Constructing the Argument

As an Artificial Lawyer reader, you are most likely already a convert to the long-term benefits of AI-powered automation. However, amongst colleagues you may be facing subtle resistance, budget inertia, or outright denial from those who view AI as a passing IT upgrade rather than a fundamental restructuring of business.

To sell the argument internally, leadership needs to frame AI strategy around human preservation and structural resilience, not just procurement. Training models must move from content creation to the apprenticeship of oversight – teaching juniors how to audit adversarially: to test and challenge autonomous systems.

Our latest report, ‘AI, Human Capital and the Future of Professional Services’, provides the data, insights, and framework to help you build this business case. It provides a blunt ‘state of the nation’ appraisal of the market, mapping out how to bridge the volatility gap between technological advancement and institutional longevity.

The full report is free to download:

👉 Download the Full Report Here

About the author: James Tuke is the founder of multiple tech-related ventures, such as Treat Digital, and is now also the CEO of the AI Futures Forum, a site dedicated to in-depth discussion about AI, and writer of ‘A Short Walk in AI’. He also advises firms and senior management about their AI strategy.